Rapidly Printing with Liquid Metal

Andrew Corselli

MIT researchers have developed an additive manufacturing technique that can print rapidly with liquid metal, producing large-scale parts like table legs and chair frames in a matter of minutes.

Their technique — called liquid metal printing (LMP), which the team says is at least 10 times faster than a comparable metal additive manufacturing process — involves depositing molten aluminum along a predefined path into a bed of tiny glass beads. The aluminum quickly hardens into a 3D structure.

In a recent study, the researchers demonstrated the procedure by printing aluminum frames and parts for tables and chairs which were strong enough to withstand post-print machining. They showed how components made with LMP could be combined with high-resolution processes and additional materials to create functional furniture.

Here is an exclusive Tech Briefs interview — edited for length and clarity — with Associate Professor and Senior Author Skylar Tibbits.

Tech Briefs: What was the biggest technical challenge you faced while developing LMP?

Tibbits: Just to give you the backstory, the work comes out of the Self-Assembly Lab, at which I’m the co-director. The precursor to this one was rapid liquid printing. We developed a printing technology into gel, and we spun that out as a company. In my mind, I'm always thinking of that; this was developed after, thinking about ‘how do we go for higher temperatures?’ Because that was one limitation on the rapid liquid printing. That's kind of why we got into this.

But the moment you start to go higher temperature, you run into all sorts of crazy problems. The nozzle is where 90 percent of the problems are — the three-axis part of it and the gantry system and the powders and even the heating of the crucible and the furnace, it's all barely resolved. The nozzle is where everything falls apart and gets complicated because you're at a super-high temperature and the interface is between materials, so how do you not get cold spots and how do you not get material corrosion issues?

Flow control in the nozzle is where all of it is challenging.

Tech Briefs: Can you explain in simple terms how it works?

Tibbits: Basically, we have a furnace, and we melt aluminum. Usually, we use scrap aluminum from campus, and we melt it. So, it could be like ingots of material or beads or whatever it is. And that gets melted at 600-700 °C in a furnace, sitting in a crucible. Then, from the crucible there's a nozzle — we went through about 20 different nozzle designs — through which the material flows.

We print inside of a bed of glass beads — very fine powder — basically it feels like silk in your hands. The glass beads that receive the high temperature cool it down, but also provide a support.

Then it's a three-axis machine, so we can move with CNC in three axes and extrude out, and it happens very fast, usually under a minute. Materials rapidly cool and rapidly flow, and we can print large scale. Right now, the machine's 2-foot-by-4-foot, but the gantry itself could be 5-by-10 or it could be full-room scale. The gantry is really standard; it's just the furnace and the nozzle and the beads that are different.

Tech Briefs: What are the pros and cons of this technique?

Tibbits: In the beginning, our main goal was ‘how do we print large and fast?’ There’s a handful of metal printing technologies now — most of them are trying to chase after higher and higher precision. That's because of high-quality, high-value parts in aviation, automotive, engine components, etc.

So, sintered metal powders or various other technologies give you high resolution, but they're super slow and super small scale. We wanted to do the opposite. Coming from a background in architecture and working with a lot of furniture companies, we look at large-scale things that need low resolution.

We're never going to be sintering; lasered, sintered powders are always going to have higher resolution. But we're under a minute, where they're under a day. It would take many hours to make this part; and we're meters and they're centimeters. Speed and scale are our main advantages; resolution is our disadvantage.

It's totally recyclable, which is also nice. Everything we print, we can melt back down and print again. So that has a nice advantage. And then, lastly, the way that we think we can overcome the precision limitations is that you can post-machine it. We basically print out roughly what you need, and then you can get all the precision you need afterwards by machining it.

The researcher, [Lead Author] Zain [Karsan], was the master student; this was his thesis. He's now at ETH doing his PhD, but in one of the demonstrations he did a three-axis CNC machine to show that you could make as precise a machine as you need. All the parts are printed, but the details and the connections, that's all machined. So, you can still get that precision.

Tech Briefs: Do you have any other future research, work, etc. on the horizon?

Tibbits: I think we'd love to continue to push this at the larger scale — furniture and up. That’s our goal.

We've done two rounds of this machine. Phase one, we were printing with pewter, so lower temperature and we had a much smaller nozzle, a little bit better flow control, and the resolution was better. Phase two, we scaled up and went to aluminum — a lot harder because it's higher temperature, but also, we went to much larger cross sections in much larger parts.

So, phase three, I think we need to redesign the nozzle construction and really get flow control. That’s our main focus now. I think we demonstrated the things we were trying to demonstrate with aluminum, but now it's about how do we make this a viable technology that could be used in a factory setting or be spun out? The way that we do that is flow control and repeatability. Those are the main issues we need to demonstrate.

Tech Briefs: Do you have any advice for engineers aiming to bring their ideas to fruition?

Tibbits: Work hard, fail a lot, keep trying, don’t give up, and have amazing people around you. We're a research lab, so our whole goal is to go from impossible to possible. So, we're allowed to fail; we're not limited by profitability or customer demand or economy.

It takes time, and you fail often, and it costs a lot of money, and it's frustrating. It’s been years of working on this and it's still super hard — it doesn't get any easier. So, I think it's just finding the right situation to be able to put in that hard work to actually innovate. It doesn't come easy.

---